study of the primary decomposition of coal by infrared
TRANSCRIPT
A STUDY OF THE PRIMARY DECOMPOSITION OF COAL BY INFRARED
SPECTROPHOTOMETRY AND BY CHLOROFORM EXTRACTION
Raymond Virgil Smith
A thesis submitted to the faculty of the Uni"versity of Utah in partial fulfillment of the requirements for the degree of
Master of Science
Department of Fuel Technology LIBRARY
UNIVERSITY OF UTAH
University of Utah August, 1957
A STUDY OF THE PRIMARY DECOMPOSITION OF COAL BY INFRARED
SPEC TROPHOTOMETRY AND BY CHLOROFORM EXTRACTION
by
Raymond. Virgil Smith
A thesis submitted to the faculty of the University of Utah in partial fulfillment of the requirements for the degree of
Master of Science
Department of Fuel Technology
University of Utah
August, 1957
,/ LIBRARY !JNIVERSITY OF. UTAH
This Thesis for the Master of Science degree
by
Raymond Virgil Smith
has been approved by
� �� (�. � ReadeJSUi)ei'Visory Committee
ABSTRACT
For the differential I-R technique described in this paper,
two specimens of coal were prepared for runs in the infrared double
beam spectrometer. One specimen was prepared for the coal as
received from the mine and this was placed in the reference beam.
The other sample was prepared from coal which had been heated in
a pressure tight container to the softening temperature. This was
placed in the sample beam. The differential infrared spectrometer
pattern thus obtained enables one to observe the changes in the
infrared range.
In the second phase of these tests the coal was heated to
temperatures in the plastic range and then extracted with chloroform.
The extract yield was run differentially versus the untreated coal
in the infrared spectrophotometer. These tests indicated different
band intensities than the untreated coal and also revealed some
absorbtion bands which did not occur in the original coal or in the
residue extract from the absorbtion process.
The extract yield data was also used for a kinetic study of
the coal!s primary decomposition. Activation energies thus obtained
for the solid to plastic step of the reaction appear to be of the
general order of magnitude of 20 to 30 k.cal/mole*
ii
ABSTRACT
For the differential I-R technique described in this paper"
two specimens of coal were prepared for runs in the infrared double
beam spectrometer. One specimen was prepared for the coal as
recei ved from the mine and this was placed in the reference beam.
The other sample was prepared from coal which had been heated in
a pressure tight container to the softening temperature. This was
placed in the sample beam. The differential infrared spectrometer
pattern thus obtained enables one to observe the changes in the
infrared range.
In the second phase of these tests the coal was heated to
temperatures in the plastic range and. then extracted with chloroform.
The extract yield was run differentially versus the untreated coal
in the infrared spectrophotometer. These tests indicated different
band intensities than the untreated coal and also revealed some
absorbtion bands which did not occur in the original coal or in the
residue extract from the absorbtion process.
The extract yield data was also used for a kinetic study of
the coal r s primary de compo si tion. Activation energies thus obtained
for the solid to plastic step of the reaction appear to be of the
general order of magnitude of 20 to 30 k.cal/mole.
i1
Acknowledgements
The author wishes to express his thanks to Dr« George
Richard Hill, Department of Fuel Technology and Dr. Milton E.
Wadsworth for their interest and help throughout this stucty-*
This research was supported in part by a J. L. Dougan
Research Fellowship in Fuel Technology and the Atomic Energy
Commission under Contract Number AT ( 1 1 - 1 ) -82, Project Number 1«
Their interest and support was greatly appreciated*
Thanks are also expressed to the U» S* Steel Corporation,
Columbia Division who supplied the coal samples for the tests.
iii
Acknowledgements
The author wishes to express his thanks to Dro George
Richard Hill, Department of Fuel Technology and Dr. Milton E.
Wadsworth f'or their interest and help throughout this study.
This research was supported in part by a J. L. Dougan
Research Fellowship in Fuel Technology and the Atomic Energy
Commission under Contract Number AT (11-1) -82, Project Number 10
Their interest and. support was greatly appreciated.
Thanks are also expressed to the U. S. Steel Corporation,
Columbia Division who supplied the coal samples for the tests.
iii
TABLE OF CONTENTS
I. INTRODUCTION 1
II. INFRARED SPECTROPHOTOMETRY STUDY OF THE COAL HEATED TO THE PLASTIC STAGE 8
III. INFRARED SPECTROPHOTOMETRY STUDY OF THE CHLOROFORM EXTRACTION OF HEATED COALS l£
IV. KINETIC STUDY OF THE FORMATION OF A CHLOROFORM
SOLUBLE MATERIAL 19
V. CONCLUSIONS • . . • 23
VI. RECOMMENDATIONS 2$
APPENDICES. . • 26
BIBLIOGRAPHY 33
iv
TABLE OF CONTENTS
I. INTRODUCTION. • • • • 0 • • • • • • • 1
TI. INFRARED SPECTROPHOTOMETRIC STUDY OF THE COAL HEATED TO THE PLASTIC STAGE. • • • • • • 0 • 8
III. INFRARED SPECTROPHOTOMETRIC STUDY OF THE CHLOROFORM EXTRACTION OF HEATED COALS 0 • • • • 15
IV. KINETIC STUDY OF THE FORHATION OF A CHLOROFORM SOLUBLE MATERIAL • • • • • • • • • • • o 19
v. CONCLUSIONS • • • • • 0 • • • • • • • • 23
VI. RECOM}lENDATIONS. • • • • • • • • 0 • • • 25
APPENDICES. • • • • • • • • • 0 • • • 0 0 • 26
BIBLIOGRAPHY • • • • • • • • • • • • • • • • 33
iv
I. INTRODUCTION
The primary decomposition of coal is closely related to the
quality of coke the coal will ultimately produce.
Coke and Coke Quality 9
Fuels and Combustion Handbook defines coke as a "•••fused,
cellular, porus structure that remains after free moisture and the
major portion of the volatile matter have been distilled from
bituminous coal and other carbonaceous material by the application
of heat in the absence of air or in limited supply*" Chemistry of
Coal utilization10 states that about 1$% of the total annual coal
production has been coked for a number of years.
The required quality of the finished coke for blast furnace
use is very loosely defined. Fuels and Combustion Handbook^ states
"The real answer is to be had only from the blast furnace."
Chemistry of Goal Utilization adds, "The correct evaluation of
blast furnace coke is s till an unsolved problem."
Two tests which determine the strength and therefore the
acceptability of the coke are the Drop Shatter (A.S.T.M. D1U1-U8) and
Tumbler (A.S.T.M. D29U-50) Tests •
In the Shatter Test, 50 pounds of coke all over 2 inches in
size are placed in a box and dropped four times from a height of 6 1
I. INTRODUCTION
The primary d.ecomposi tion of coal is closely related to the
quality of coke the coal will ultimately produce.
Coke and Coke Qual! ty
Fuels and Combustion Handbook9 defines coke as a n •• ofused,
cellular" porus structure that remains after free moisture and the
major portion of the volatile matter have been distilled from
bituminous coal and other carbonaceous material by the application
of heat in the absence of air or in limited supply." Chemistr,r of
Coal utilizationlO states that about 15% of the total annual coal
production has been coked for a number of years.
The required quality of the finished coke for blast furnace
use is very loosely defined. Fuels and. Combustion Handbook9 states
"The real answer is to be had only from the blast furnace."
Chemistry of Coal Utilization adds, "The correct evaluation of
blast i'urnace coke is s till an unsolved problem."
Two tests which determine the strength and therefore the
acceptability of the coke are the Drop Shatter (A.S.T.M. D14l-48) and
Tumbler (A.S.T.M. D294-50) fests •
In the Shatter Test, 50 pounds of coke allover 2 inches in
size are placed in a box and dropped four times from a height of 6
1
2
feet on a steel plate. The resulting product is sieved on 2, if, and
1 inch screens. The Shatter Index is the amount remaining on a
given size screen.
In the Tumbler Tests coke is placed in a drum and rotated
at a specified speed for a fixed length of time. There is statis
tical correlation between coke quality and the Shatter and Drum Tests©
The size as well as the strength of the coke is an important
factor in the blast furnace. A uniform size of coke seems to produce
the greatest void value and the best ratio of coke consumption per
ton of iron produced.
There are many other tests for coke quality which include
determinations of reactivity, density, combustability, absorbtivity,
electrical conductivity and compression strength.
Tests to Determine Coking Quality
Determination of the rank of the coal proves to be a rough
measure of the coal's coking quality. In general, the low volatile,
bituminous coal produces the best coke. Most papers in this search
indicate that coal with a carbon content in the Q0% to 90% range
produce the best coke.
Chemistry of Coal Utilization3*0 lists nine types of tests to
determine the coking quality. Those most often referred to in the
literature are the Free Swelling, (A.S.T.M. D?20-l;6) Giesler
Plastometer and Dilatometer Tests*
Two types of Free Swelling Tests are Agglutinating and Agglomerating. In the Agglutenating Test, one gram of coal with
2
feet on a steel plate. The resulting product is sie-ved on 2, Ii, and
1 inch screens. The Shatter Index is the amount remaining on a
given size screen.
In the Tumbler Tests coke is placed in a drum and rotated
at a specified speed for a fixed length of time. There is statis
tical correlation between coke quality and the Shatter and Drum Tests.
The size as well as the strength of the coke is an important
factor in the blast .furnace 0 A uniform size of coke seems to produce
the greatest void value and the best ratio of coke consumption per
ton of iron produced.
There are many other tests for coke quality which include
determinations of reactivity, density, combustability, absorbtivity"
electrical conductivity and compression strength.
Tests to Determine Cokin;; Quality
Determination of the rank of the coal proves to be a rough
mea'5Ure of the coal's coking quality. In gereral" the low volatile,
bituminous coal produces the best coke. Most papers in this search
indicate that coal with a carbon content in the 80% to 90% range
produce the best coke.
Chemistry of Coal UtilizationlO lists nine types of tests to
determine the coking quality. Those most often referred to in the
literature are the Free Swelling, (A.S.T.M. D720-46) Giesler
Plastameter and Dilatometer Tests.
Two types of Free Swelling Tests are Agglutinating and
Agglomerating. In the Agglutenating Test, one gram of coal with
3
varying amounts of inert material is carbonized at 950 degrees C. The
recorded results are the maximum weight of inert material per gram
of coal that will produce a coke button which will sustain a 5>00 g.
weight without crushing. In the Agglomerating Test, one gram of
coal is heated at 9^0 degrees C. The tendency to swell and the
ability to support a $00 g© weight are observed.
The Dilatometer testing method uses a piston on the coal,
free to move as the coal swells on heating. The vertical displacement
of the piston is recorded.
In the Giesler Plastometer Test a constant torque is applied
on wire-like fingers in the coal throughout a temperature programmed
heating process. Dial divisions of rotation of the torqued fingers
in the coal are recorded versus temperature.
Data and Theories on the Coking Mechanism
The key to coking seems to be in the fluid behavior of the
coal as it is heated. This pointed up by the Free Swelling and
Plastometer Tests and, more recently, by the work of Dryden and
Panchurst.1 They discovered that a chloroform soluble product is
formed near the softening temperature. They proved that this ex
tracted portion was an important part of the plastic stage by three
tests. In the first test a coking coal would not swell on heating
after the extracted portion had been removed. Attempts to extract
material from a non-coking coal did not result in a good yield of
soluble product and the yield did not rise as the softening
3
varying amounts of inert material :i,.s carbonized at 950 degrees C. The
recorded results are the maximum weight of inert material per gram
of coal that will produce a coke button which will sustain a 500 g.
weight without crushing. In the Agglomerating Test, one gram of
coal is heated at 950 degrees C. The tendency to swell and the
ability to support a 500 go rleight are observed.
The DilatOllleter testing method uses a piston on the coal,
free to move as the coal swells on heating. The vertical displacement
of the piston is recorded.
In the Giesler Plastometer Test a constant torque is applied
on wire-like fingers in the coal throughout a temperature programmed
heating process. Dial divisions of rotation of the torqued fingers
in the coal are recorded versus temperature.
Data and Theories on the Coking Mechanism.
The key to coking seems to be in the fluid behavior of the
coal as it is heated. This pointed up by the Free Swelling and
Plastometer Tests and, more recently, by the WOrk of Dryden and
Panchurst. l They discovered that a chloroform soluble product is
fonned near the softening temperature. They proved that this ex
tracted portion was an important part of the plastic 'stage by three
tests. In the first test a coldng coal would not swell on heating
after the extracted portion had been removed. Attempts to extract
material from a non-coking coal did not result in a good yield of
soluble product and the yield did not rise as the softening
temperatures were reached as it did in the coking coal* In the
third tests, adding the extract yield to a non-coking coal produced
softening and swelling properties in it which it did not possess
prior to the addition* Further tests on the extracted portion
indicated that the extract was more fluid on heating than the origi
nal coal* They found that the extract contained more hydrogen and
less oxygen than the original coal* They also stated that an
infrared spectra showed that both the extract and the residue
resembled the original coal* The yield of extract was found to
drop off rather sharply if softening temperatures were held for
longer periods of time during t&e heating process*
There is a good deal of infrared spectrophotometry data on
coal* Some of the points brought out by previous studies follow •
1* Four papers^' 6* 7* 1 1 all report heavy OH bands (2.7 to 3*1
microns) in high volatile matter, poor coking coal* This bond disap
pears on heating. In general, this bond does not appear in the
spectra of higher ranked coals which have good coking characteristics* 2 7 11 12
Several papers >*> ' all suggest that the low rank, high volatile
matter coal!s inability to coke is because of the strong linkages of
these OH groups* J. K. Brown^ however, states "The simple explanation
that strong intermolecular forces associated with hydrogen-bonding
restricted the ability of the weakly coking coal to become plastic
and swell is not satisfactory and is not in accord with a number of
facts known about the coking process (e.g* the destruction of coking
power by solvent extraction)*11
4
temperatures were reached as it did in the coking coal. In the
third tests, adding the extract yield to a non-coking coal produced
softening and swelling properties in it which it did not possess
prior to the addition. Further tests on the extracted portion
indicated that the extract was more fluid on heating than the origi
nal coal. They found that the extract contained more hydrogen and
less o:xygen than the original coal. They also stated that an
infrared spectra showed that both the extract and the residue
resembled the original coal. The yield of extract was found to
drop off rather sharply if softening temperatures were held for
longer periods of time during the heating process.
There is a good deal of infrared spectrophotometr;ic data on
coal. Some of the points brought out by previous studies follow •
1. Four papers~,6,7,ll a1l report heavy OH bands (2.7 to 3.1
microns) in high volatile matter, poor coking coal. This 'bond disap
pears on heating. In general, this bond does not appear in the
spectra of higher ranked coals which have good coting characteristics.
Several papers2,7,ll,12 all suggest that the low rank, high volatile
matter coalts inability to coke is because of the strong linkages of
these OH groups. J. K. Pirmm6 however, states "The simple explanation
that strong intermolecular forces associated with hydrogen-bonding
restricted the ability of the weakly coking coal to become plastic
and swell is not satisfactory and is not in accord with a nUlllber of
facts known about the coking process (eg. the destruction of coking
power by solvent extraction)."
2* H. H. Storch indicates weathered and oxidized coals
show a strong carbonyl (5*87 microns) band absorption. Kinkby,
Lakey, and Sareant1^ find this a strong absorption band in low rank
coals*
3* There is some confusion on the 6*19 micron band* H* H.
Storch points out that it is hard to separate from the hydrogen
bonded OH at 6*1 microns and 6*20 microns* He concludes, however
that the absorption at 6.19 microns is a result of phenoxyl or
quinoidal compounds* J. K. Brown^ found the 6*19 micron absorption
in all coals except anthracite. ?He identified the band with an
aromatic ring frequency or with a carbonyl group* He does not
believe this absorption indicates a phenolic group because the
intensity of the absorption does not decrease with the carbon content
in the coal.
lu Three papers^'7*11 report a rather strong absorption in
the 8 micron region. Two of these authors^'11 state this is the
result of aromatic oxygenated compounds* One of the papers suggests
that this absorption indicates the presence of polycyclic quinone
compounds. They find this absorption band quite stable for coal
ranks 78$ to 89$ carbon*
5* Fridel and Queiser^ report that kstolinite or an aromatic
ether represent the 9*67 micron absorption band*
6. Several papers report that high background absorption
increases with the rank of the coal and this background may be
5
20 H. H. Storchll indicates weathered and oxidized coals
show a strong carbonyl (5.87 microns) band absorption. Kinkby,
Lakey I and Sareantl6 find this a strong absorption band in low rank
coals.
3. There is some confUsion on the 6.19 micron band. H. H.
storch points out that it is hard to separate from the hydrogen
bonded OH at 6.1 microns and 6.20 microns. He concludes, however
that the absorption at 6.19 microns is a result of phenoxy1 or
quinoida1 compounds. J. K. Brown6 found the 6.19 micron absorption
in all coals except anthracite. '?He identified the band with an
aromatic ring frequency or with a carbonyl group. He does not
believe this absorption indicates a phenolic group because the
intensity of the absorption does not decrease with the carbon content
in the coal.
4. Three papers6,7,llreport a rather strong absorption in
the 8 micron region. Two of these authors6,ll state this is the
result of aromatic oxygenated compoundso One of the papers suggests
that this absorption indicates the presence of polycyclic quinone
compounds. They find this absorption band quite stable for coal
ranks 78% to 89% carbon.
541 Fridel and Queiser5 report that kaolinite or an aromatic
ether represent the 9.67 micron absorption band.
6. Several papers report that high background absorption
increases with the rank of the coal and this background may be
6
caused by pi electrons^ in condensed aromatic rings or by scattering.7
J. Ko Brown^ used infrared spectrophotometry to study a
weakly caking coal and a coking coal heated to various temperatures.
As a result of this study he suggested that the removal of the
aliphatic hydrocarbons leaves unsatisfied edge valences which are
satisfied by C-C cross links. This reaction, he believes, continues
above degrees C. (above which the high background opaqued his
spectra) with a further loss of hydrogen from the ring systems. 7
Brown and Hirsch' found by use of X-Ray diffraction techniques
that the number of condensed aromatic rings goes up rapidly for
coals with about 87$ carbon content. They report that 1$% of the
carbon in a coal containing a total of 85$ carbon is in condensed
aromatic rings.
A. Whitacker"""̂ using the data produced by the X-Ray technique,
studied the fringe groups -which satisfy the edge valences of the
carbon rings. From the amount of unorganized material and from his
computed number of edge valences he concluded that there is barely
enough unorganized material to go around to satisfy the edge valences
unless there are considerable direct linkages. His work indicates
that fringe structures, for the most part, must be composed of
rather simple groups.
Fitzgerald1^ made a kinetic study of the plastic zone
reaction, evaluating the reaction velocity from the Giesler Plasto-
meter data. He points up the fact that fluidity is a function of
6
caused by pi electrons5 in condensed aromatic rings or by scattering.7
J. Ko F5rawn6 used infrared spectrophotometry to study a
weakly caking coal and a coking coal heated to various temperatures.
As a result of tlUs study' he suggested that the removal of the
aliphatic hydrocarbons leaves unsatisfied edge valences which are
satisfied by C-C cross links. This reaction, he believes, continues
above 550 degrees C. (above which the high background opaqued his
spectra) with a further loss of hydrogen from the ring systems.
Brown and Hirsch 7 found by use of X-Ray diffraction techniques
that the number of condensed aromatic rings goes up rapidly for
coals with about 87% carbon content. They report that 75% of the
carbon in a coal containjng a total of 85% carbon is in condensed
aromatic rings.
A. 'Whitacker12 using the data produced by the X-Ray technique,
studied the fringe groups which satisfy the edge valences of the
carbon rings. From. the amount of unorganized material and from his
computed number of edge valences he concluded that there is barely
enough unorganized material to go around to satisfy the edge valences
unless there are considerable direct linkages. His work indicates
that fringe structures, for the most part, must be composed of
rather simple groups.
Fitzgerald13 made a kinetic study of the plastic zone
reaction, evaluating the reaction velocity from the Giesler Plasto
meter data. He points up the fact that fluidity is a function of
7
both temperature and time because at a constant temperature the
fluidity goes to a maximum and then falls off. He observed that
this indicates a decomposition reaction, first to a fluid, and then
a decomposition of the fluid to solid and volatile products* He
used the Arrhenius rate equation and found the activation energy
for several coals to be 5>0 k. cal/mole© He found a reasonably
good correlation between reaction velocity and the lj inch Shatter
Index, and suggested that the use of maximum fluidity with the
reaction velocity should produce better resultso 7 11
Two authors'>^ comment that the strength of the coal
(Young!s Modulus, grindability, hardness, and viscosity) are at a
minimum in the good coking coalsc
7
both temperature and time because at a constant temperature the
fluidity goes to a max:ilmlm and. then fa1ls of:!. He observed that
this indicates a decomposition reaction, first to a fluid, and then
a decomposition of the fluid to solid and volatile products. He
used the Arrhenius rate equation and found the activation energy
for several coals to be 50 k. cal/moleo He found a reasonably
good correlation between reaction velocity and the It inch Shatter
:Index, and suggested that the use of maximum fluidity with the
reaction velocity should produce better resultso
Two authors7,ll comment that the strength of the coal
(Young's Modulus, grindability, hardness, and viscosity) are at a
minimum in the good coking eoalso
II. INFRARED SPECTROPHOTOMETRY STUDY OF COAL HEATED TO THE PLASTIC STAGE
Procedure in the Preparation of Samples
Coal specimens were prepared for evaluation by means of
infrared spectrophotometry. The samples were ground to a very fine
powder and then thoroughly mixed in carefully weighed proportions
with potassium iodide (usually 6mg. of coal per gram of KI)« A measured
quantity of this mixture was then pressed in an evacuated die into
plates for examination. The reliability of this technique has been 2 Q h
well established in other systems.'-^'^ Infrared specta were
obtained using a Perkin Elmer Model 21 double beam recording
spectrophotometer.
In the first tests the infrared absorbtion spectrum was
obtained for three coals with a high fluidity (Giesler Plastometer
dial divisions greater than 100) and for three coals with poor fluid
properties (Geisler Plastometer dial divisions less than three).
Most of the spectral assignments have been made in previous studies.
Table II lists the spectral assignments used in this study.
Discussion of Results
For the 2*7 to 3*10 micron band the non-fluid coals showed
stronger absorbtion indicating more OH groups in these coals* This
8
110 INFRARED SPECTROPHOTOMETRIC STUDY OF COAL
HEATED TO THE PLASTIC STAGE
Procedure in the Preparation of Samples
Coal specimens were prepared for evaluation by' means of
infrared spectrophotometry. The samples were ground to a very fine
powder and then thoroughly mixed in carefully weighed proportions
with potassium iodide (usually" 6mg. of coal per gram of KI)o A measured
quantity of this mixture was then pressed in an evacuated die into
plates for examination. The reliability of this technique has been
well established in other systems?,3,4 Inf'rared specta were
obtained using a Perkin Elmer Model 21 double beam recording
spectrophotometer.
In the first tests the infrared absorbtion spectrum was
obtained for three coals with a high fluidity (Giesler Plastometer
dial divisions greater than 100) and for three coals with poor fluid
properties (Geisler Plastometer dial divisions less than three).
Most of the spectral assignments have been made in previOUS studies.
Table II lists the spectral assignments used in this study.
Discussion of Results
For the 2.7 to 3.10 micron band the non-fluid coals showed
stronger absorbtion indicating more OH groups in these coalso This
8
9
was in agreement with the findings of previous investigatorsJ>*^'t*^m
Two of the fluid coals showed some carbonyl bonds (5*87 microns)*
None of the other coals absorbed in this region* All of the coals
absorbed strongly at the 6*19 micron band again in agreement with
the previous investigations^ In the 8*0 micron region all coals
showed weak absorbtion* None of the coals tested absorbed at the
9*05> micron ether band* The fluid coals indicated some absorbtion
in the 9*67 micron ether region* This study was limited to the
shorter wave lengths* and the more specifically assigned bands*
The results of this series of tests indicated that the
spectra of the coals of the Rocky Mountain Region are similar to the
spectra of the coals of the other parts of the world* The spectra
also point up the fact that while some indication of a coal!s fluid
properties may be obtained from the infrared pattern, these data
are not sufficient to evaluate the fluid properties of the coal*
Procedure for Differential I-R Examination of Heated Coal
The coal used for the second tests in this series was a
blending coal used in coke ovens* The physical properties of this
coal are shown in Table I. One sample was prepared from the coal as
received from the mine* The other sample of coal was heated in a
pressure tight container to the fluid ranger this required approximately
20 minutes* The temperature in the fluid range was held for periods
9
was in agreement with the findings of previous investigators.5,6,7,l1
Two of the fluid coals showed some carbonyl bonds (5.87 microns).
None of the other coals absor1Ded in this region. All of the coals
absorbed strongly at the 6.19 micron band again in agreement with
the previous investigations. In the 8.0 micron region all coals
showed weak absorbtion. None of the coals tested absorbed at the
9.05 micron ether band. The fluid coals indicated some absorbtion
in the 9.67 micron ether region. This study was limited to the
shorter wave lengths, and the more specifically assigned bands.
The results of this series orliests indicated that the
spectra of the coals of the Rocky Mountain Region are similar to the
spectra of the coals of the other parts of the world. The spectra
also point up the fact that while some indication of a coal I s fluid
properties may be obtained from the infrared pattern, these data
are not sufficient to evaluate the fluid properties of the coal.
Procedure for Differential I-R Examination
of Rea ted Coal
The coal used for the second. tests in this series was a
blending coal used in coke ovens. The physical properties of this
coal are shown in Table I. One sample was prepared from the coal as
received from the mine. The other sample of coal was heated in a
pressure tight container to the fluid range: this required approximately
20 minutes. The temperature in the fluid range was held for periods
10
Table 1. Physical Properties of Coal Used in Differential Infrared Spectrophotometer Tests
Proximate Analysis Ultimate Analysis Gieseler Type Plastometer
H 20 C 75o20 Initial Soften. Temp. 3Ul°C
V.- M. 30.53 % &23 Max. Fluidity Temp. U22°C
F. C 58.17 N 2 1.63 Max. Dial Div./Min. 2,977
Ash r u 3 0 Ash 11*30 Solidification Temp. U69°C
Sulfur 0.82 S 0.82
°2 5o82
Table II. Spectral Assignments for Coal
Microns
2.7U - 2.78 Free OH stretching 2.82 - 3 .10 OH stretching, intermolecular hydrogen bonds 3.30 Aromatic CH, weak 3.1|.2 - 3«U9 Naphthenic and/or aliphatic CH bonds 5*87 C==°=0 band, weak shoulder 6.19 Very intense band; may be partly caused by a con
jugated carbonyl structure such as in quinones 6.90 CH2 groups 7.25 CH- groups 9.05 Ether band 9.67 Aromatic band, intense in aromatic ethers
10
Table 1. Physical Properties of Coal Used in Differential Infrared Spectrophotometer Tests
Proxima te Analysis Ul tima te Analysis Gieseler TYPe P1astometer
H2O 2094 C 75020 Ini tial Soften. Temp. 341°c
V. M. 30.53 ~ $.23 Max. Fluidity Temp. 422°C
F. C. 58.17 N2 1063 Max. Dial Div./Min. 2,977
Ash 11.30 Ash 11.30 Solidification Temp. 469°C
Sulfur 0.62 S 0.82
O2 5 0 82
Table II. Spectral Assigmnents for Coal
Microns
2.74 - 2.78 2.82 - 3.10 3.30 3.42 - 3.49 5.87 6.19
6.90 7.25 9.05 9.67
Free OH stretching OH stretc~ng, intermolecular hydrogen bonds Aromatic CH, weak Naphthenic and/or aliphatic CH bonds C-==O band, weak shoulder Very intense band; may be partly caused by a conjugated carborw1 structure such as in quinones CH2 groups CH":\ groups Etlier band Aromatic band, intense in aromatic ethers
1 1
of time from 10 to 30 minutes with all time lengths producing the
same general spectra. The sample was then quenched and prepared
as previously described.
The plate containing the untreated coal was placed in the
reference beam of the spectrophotometer and the plate containing
the coal which had been heated was placed in the sample beam. A
differential absorption spectra was thus obtained.
To prove the reliability of the differential spectrophotometer
method, two samples of untreated coal were prepared and one was
placed in the reference beam and the other placed in the sample
beams. The pattern obtained is shown in Figure 2 . The virtual
absence of absorption maxima substantiated the reliability of the
differential technique and the data obtained from it.
Discussion of the Results of the
Heated Coal Series
Figure 1 shows a typical spectrum obtained by the differential
technique. Since the coal sample which had been heated was placed
in the sample beam, peaks downward on the pattern indicated vibrational
spectra for bonds greater in number in the heated coal and peaks
upward indicated the presence of a greater number of bonds in the
untreated coal sample. Increases were noted in the 2.7 to 3*10
(OH stretching) micron range, 3»3 "bo 3«U micron (aromatic hydrocarbon),
7.2^ micron (hydrocarbon bending, CE,) and 9#67 micron (aromatic ethers).
11
of time from 10 to 30 minutes with all time lengths producing the
same general spectra. The sample was then quenched and prepared
as previously described.
The plate containing the untreated coal was placed in the
reference beam of the spectrophotometer and the plate containing
the coal which had been heated was placed in the sample beam. A
differential absorption spectra was thus obtained.
To prove the reliability of the differential spectrophotometer
method, two samples of untreated coal were prepared and one was
placed in the reference beam and the other placed in the sample
beams. The pattern obtained is shown in Figure 2. The virtual
absence of absorption maxima substantiated the reliability of the
differential technique and the data obtained from it.
Discussion of the Results of the
Rea ted Coal Series
Figure 1 shows a typical spectrum obtained by the differential
technique. Since the coal sample which had been heated was placed
in the sample beam, peaks downward on the pattern indicated vibrational
spectra for bonds greater in number in the heated coal and peaks
upward indicated the presence of a greater number of bonds in the
untreated coal sample. Increases were noted in the 2.7 to 3.10
(OR stretching) micron range, 3.3 to 3.4 micron (aromatic hydrocarbon),
7.25 micron (hydrocarbon bending, CH3 ) and 9.67 micron (aromatic ethers).
12
,—OH + H0-
\ / Other studies show that coal is primarily composed of condensed
aromatic rings. Experimental work discussed later in this paper and by
others^ indicates that the final coke satisfies these edge valences with
C - C linkages. This suggested mechanism is based on the theory that
heating coal is a continuous change of edge groups and edge bonds of
these rings. Some hydrogen would of course be available from the
decomposition of the CH groups. An aromatic ether could be a fluid in
the softening temperature range. The action of this fluid could provide
greater mobility for the rings which, in turn, may result in a better,
final orientation of the crystallites. This action would make a greater
number of C - C bonds and consequently a stronger coke.
The increases in the aromatic (3.3 micron - 3»U micron) bond
substantiates the general theories of the coking mechanism which
suggest changes to the aromatic and ultimately to a graphitic structure.
There does not seem to be any ready explanation for the increase in the
CH3 band. This, too, may represent a change in edge groups. The CH2
band shows a decrease. The £.87 micron (CO) and the 6.19 micron
(C«0-quininoid) bands are shown to increase; however, some of the tests
indicated a decrease. It would seem most probably that the 0*0 band
does decrease, however the possibilities of HgO also contributing to
one of these bands (6,19 micron) and of possible oxidation of the
The increases in aromatic ether bonds suggest the following mechanism,
^ \ - 0 - / H
12
The increases in aromatic ether bonds suggest the following mechanism.
+ HO-( \
Other studies show that coal is primarily composed of condensed
aromatic rings. Experimental work discussed later in this paper and by
others8 indicates that the final coke satisfies these edge valences with
C - C linkages. This suggested mechanism is based on the theory that
heating coal is a continuous change of edge groups and edge bonds of
these rings. Some hydrogen would of course be available fram the
decomposition of the CH groups. An aromatic ether could be a fluid in
the softening temperature range. The action of this fluid could provide
greater mobility for the rings which" in turn" may result in a better"
final orientation of the crystallites. This action would make a greater
number of C - C bonds and consequently a stronger coke.
The increases in the aromatic (3.3 micron - 3.4 micron) bond
substantiates the general theories of the coking mechanism which
suggest changes to the aromatic and ultimately to a graphitic structure.
There does not seem to be any ready explanation for the increase in the
CH3 band. This, too, may represent a change in edge groups. The CH2
band shows a decrease. The 5.87 micron (C=o) and the 6.19 micron
(C=O-quininoid) bands are sholm. to increase; however, some of the tests
indicated a decrease. It would seem most probably that the C=O band
does decrease" however the possibilities of H20 also contributing to
one of these bAnds (6.19 micron) and of possible oxidation of the
13
sample after removal from the container seems the best explanation
for the behavior of these bands© No attempt was made to interpret
bands of wave lengths greater than 9#67 micron except to note their
general increase, again substantiating the polymerization-graphiti-
zation theory©
I-R Spectra Examination of Coal Heated in a Closed Crucible
Other tests were run in a closed crucible, heated rather
quickly to temperatures in the fluid range. Samples were prepared
as before and were run against a blank of KE in the reference beam.
Figure 3 shows the results of these tests© Here the general decrease
in the bands assigned to typical edge groups is noted. One should
call attention, however, to the tendency for the OH and aromatic ether
groups to increase before finally decreasing at the higher temperatures©
The two techniques differ, of course, in that in the second
method the decomposition materials are allowed to escape© It is felt
that the first (pressure-tight container) method is a more powerful
tool for the study of the softening stage because it provides a means
for examining the bonds as they occurred in that stage before subsequent
decomposition©
X-Ray and I-R Examination of the Char
Although the char resulting from Free Swelling tests when
prepared for examination by infrared spectrophotometry produced a high
13
sample after removal from the container seems the be st explanation
for the behavior of these bands o No attempt was made to interpret
bands of wave lengths greater than 9.67 micron except to note their
general increase, again substantiating the polymerization-graphiti
zation theory.
I-R Spectra Examination of Coal Heated
in a Closed Crucible
other tests were run in a closed crucible, heated rather
quickly to temperatures in the fluid range. Samples were prepared
as before and were run against a blank of lIT in the reference beam.
Figure 3 shows the results of these tests. Here the general decrease
in the bands assigned to typical edge groups is noted. One should
call attention, however, to the tendency for the OH and aromatic ether
groups to increase before finally decreasing at the higher temperatureso
The two techniques differ, of course, in that in the second
method the decomposition materials are allowed to escape. It is felt
that the first (pressure-tight container) method is a more powerful
tool for the study of the softening stage because it provides a means
for examining the bonds as they occurred in that stage before subsequent
d.ecomposi tion.
X-Ray and I-R Examination of the Char
Although the char resulting from Free Swelling tests when
prepared for examination by infrared spectrophotometry produced a high
1U
background in the infrared absorbtion region, one sample was prepared
and successfully run using 3 rag* of coal per sample. The response was
good throught the range of frequencies. The absorbtion pattern
obtained from this test showed no absorbtion peaks* indicating a final
graphitic structure with C-C bonds.
The belief that coke is of a graphitic structure was sub
stantiated by powdering the char obtained from the coke buttons and
examining the sample by means of X-Ray diffraction. The coke button
powders showed the same general X-Ray reinforcement peaks as graphite.
These X-Ray diffraction patterns were also used to calculate the
crystallite size. These calculations showed the crystallite size for
the fluid coal to be approximately 1*5 times greater than that for
the non-fluid coal.
background in the infrared absorbtion region, one sample was prepared
and success.ful1y run using 3 mg. of coal per sample. The response was
good throught the range of frequencies. The absorbtion pattern
obtained from this test showed no absorbtion peaks,indicating a final
graphitic structure with C-C bonds.
The belief that coke is of a graphitic structure was sub
stantiated by powdering the char obtained from the coke buttons and
examining the sample by means of X-Ray diffraction. The coke button
powders showed the same general X-Ray reinforcement peaks as graphite.
These X-Ray diffraction patterns were also used to calculate the
crystallite si~e. These calculations showed the crystallite size for
the fluid coal to be approxima. te1y 1.5 time s greater than that for
the non-fluid coal.
III. INFRARED SPECTROPHOTOMETRY STUDY OF THE
CHLOROFORM EXTRACTION OF HEATED COALS
Procedure
The coal used Tor these tests was again a blending coal used
in coke ovens* Some physical properties of this coal are shown in
Table I. The coal was heated in a pressure tight container to a
temperature of 370°C in approximately 25 minutes with a period of
10 minutes at the final temperature. The sample was quenched and then
extracted in a soxhlet apparatus with chloroform. The chloroform
was then allowed to evaporate at room temperature leaving the extract
yield. This procedure was similar to that of Dryden and Panchurst
except that in these tests a pressure tight container was used as
opposed to the unsealed quartz container used by Dryden and Panchurst0
It was felt that by using the closed, higher pressure system, a more
plastic state would result and that this state would remain for a
longer period of time during heating.
Coal specimens were prepared for evaluation by means of infra
red spectrophotometry by the technique described in previous tests.
Infrared spectra were obtained using a Perkin Elmer Model 21 double
beam recording spectrophotometer. Differential spectra were obtained
by placing one sample in the reference beam and another specimen in
the sample beam. As in the previous differential tests, peaks
15
III. INFRARED SPECTROPHOTOMETRIC STUDY OF THE
CHLOROFORM EXTRACTION OF HEATED COALS
Procedure
The coal used ror these tests was again a blending coal used
in coke ovens. Some physical properties of this coal are shown in
Table I. The coal was heated in a pressure tight container to a
temperature of 3700 C in approximately 25 minutes with a period of
10 minutes at the final temperature.. The sample was quenched and then
extracted in a soxhlet apparatus with chloroform. The chloroform
was then allowed to evaporate at room temperature leaving the extract
yield. This procedure was similar to that of Dryden and Panchurst
except that in these tests a pressure tight container was used as
opposed to the unsealed quartz container used by Dryden and Panchursto
It was felt that by using the closedJ higher pressure systemJ a more
plastic state would result and that this state would remain for a
longer period of time during heating.
Coal specimens were prepared for evaluation by means of infra
red spectrophotometry by the technique described in previous tests.
Infrared spectra were obtained using a Pertin Elmer Model 21 double
beam recording spectrophotometer. Differential spectra were obtained
by placing one sample in the reference beam and another specimen in
the sample beam. As in the previous dirferential tests" peaks
15
16
downward on the pattern indicated vibrational spectra for bonds
greater in number in the specimen in the sample beam, and peaks
upward indicated the presence of a greater number of bonds in the
sample in the reference beam, Figures U,f?,6, and 7 show some of
the spectra obtained from these tests.
Discussion of the Spectra
Table II lists the spectral assignments used in this study.
The spectra obtained for the extract residue Figure h is similar to
the spectra of the untreated coal, Figure 1*
1* The differential spectra of Figure 3 shows in the OH
stretching region the extract contains less free OH and more inter-
molecular hydrogen-bonded OH than the untreated coalc
2 . Absorption in the 3©3 micron and 3»U2-3»li9 micron hydrocarbon
bands shows greater quantities of this material in the extract.
3 . In the 5.87 micron and 6.19 micron carbonyl region, the
5.87 band shows an increase in the extract. The 6.19 band shows a
rather indefinite change with the suggestion of some new bands in
that region. The apparent bands at 6.03 and 6.20 microns probably
result from absorption by OH bonds.
lw The 6.90 micron and the 7*2$ micron bands (CH2 and CH^)
show greater strength in the extract as would be expected from the
3.3 micron and 3«U2-3«29 micron bands and from previous work^ which
show that the extract contains more hydrogen than the untreated coal©
16
downward on the pattern indicated vibrational spectra for bonds
greater in number in the specimen in the sample beam, and peaks
upward indicated the presence of a greater number of bonds in the
sample in the reference beam. Figures 4,5,6, and 7 show some of
the spectra obtained from these tests.
Discussion of the Spectra
Table II lists the spectral assignments used in this study.
The spectra obtained for the extract residue Figure 4 is similar to
the spectra of the untreated coal, Figure 1.
1. The differential spectra of Figure 3 shows in the OR
stretching region the extract contains less free OH and more inter
molecular hydrogen-bonded OR than the untreated coal.
2. Absorption in the 3.3 micron and 3.42-3.49 micron hydrocarbon
bands shows greater quantities of this material in the extract.
3. In the 5.87 micron and 6.19 micron carbonyl region, the
5.87 band shows an increase in the extract. The 6.19 band shows a
rather indefinite change with the suggestion of some new bands in
that region. The apparent bands at 6.03 and 6.20 microns probably
result from absorption by OR bonds.
4. The 6.90 micron and the 7.25 micron bands (CH2 and CR3)
show greater strength in the extract as would be expected from the
3.3 micron and 3.42-3.29 micron bands and from previous workl which
show that the extract contains more hydrogen than the untreated coal.
17 5* The 9.05 micron and 9*67 micron ether bands indicate
the extract contains more of the 9•Of? vibrational absorption bands
and less of the 9*67 micron bands.
6. The extract also absorbs at four frequencies for which
there is no absorption in the untreated coal or in the residue.
These bands are at 6.6, 7*k3? &»h79 and 10.55 microns.
The 6.6 micron region most probably represents an aromatic
structure created by the softening reaction. The 7»U3 micron band
may represent an OH group for which there have been indications
in the 3 micron and 6.19 micron region. In terms of what has been
previously observed, perhaps CH^OH would be the most probable
identification for this band. A PhCHO compound could account for
the absorption in the 8.U7 and 19©55 micron wave lengths.
Taking a broad look at the extraction patterns there are
many indications of reactions of increases in the molecular OH
groups, and in the ether groups. This could be explained by a
reaction in which OH groups satisfy an edge group valence early in the
reaction finally yielding to the single bonded oxygen of the ether
group with H2O leaving the system.
It may also be noted that most of the bands are increased in
intensity in the extract. It seems reasonable to assume these bands
are for the most part edge groups on carbon ring structures. This
would indicate the extract and, in turn, the fluid portion contain
more of these edge groups. This suggests that the extract, probably
17
5. The 9.05 micron and 9.67 micron ether bands indicate
the extract contains more of the 9.05 vibrational absorption bands
and less of the.9.67 micron bands.
6. The extract also absorbs at four frequencies for which
there is no absorption in the 1.Ultreated coal or in the residue.
These bands are at 6.6, 7.43, 8.47, and 10.55 microns.
The 6.6 micron region most probably represents an aromatic
structure created by the softening reaction. The 7.43 micron band
may represent an OR group for which there have been indications
in the 3 micron and 6.19 micron region. In terms of what has been
previously observed, perhaps CH20R would be the most probable
identification for this band. A PhCHO compound could account for
the absorption in the 8.47 and 19.55 micron wave lengthso
Taking a broad look at the extraction patterns there are
many indications of reactions of increases in the molecular OR
groups, and in the ether groups. This could be explained by a
reaction in which OR groups satis~ an edge group valence early in the
reaction finally yielding to the single bonded oxygen of the ether
group with R20 leaving the system.
It may also be noted that most of the bands are increased in
intensity in the extract. It seems reasonable to assume these bands
are for the most part edge groups on carbon ring structures. This
would indicate the extract and, in turn, the fluid portion contain
more of these edge groups. This suggests that the extract, probably
18
the more significant portion of the plastic component,contains
the lower molecular weight compounds. This group of molecules
would have more edge valences to be satisfied, thus explaining the
additional bonds indicated in the extract.
18
the more significant portion of the plastic component, contains
the lower molecular weight compounds. This group of molecules
would have more edge valences to be satisfied, thus explaining the
additional bonds indicated in the extract.
IV. KINETIC STUDY OF THE FORMATION OF A
CHLOROFORM SOLUBLE MATERIAL
Theory
The following reaction mechanism is proposed for the
primary decomposition of a coal with fluid properties*
A P (1) P — SC + G^ 2)
This mechanism is similar to the mechanism proposed by Chermin and
van Krevelen^. A, represents the untreated coal; P. the plastic
stage; SC, the semi-coke and G]_, the low temperature gas products
from the primary decomposition*
In these tests chloroform extraction was used to determine
a value proportional to the amount of P produced* The relationship
between the plastic stage and the extract yield seems to be well
established by Dryden and Panchurst^ who showed that the residue does
not produce a fluid stage when the extract has been removed*
Procedure
The coal was heated in a pressure tight container as in
previous tests* In order to get reproducibility it was found that
19
IV. KINETIC STUDY OF THE FORMATION OF A
CHLOROFORM SOLUBLE MATERIAL
Theory
The following reaction mechanism is proposed for the
primary decomposition of a coal with fluid properties.
A --. P---
P (1) SC + G
I (2)
This mechanism is similar to the mechanism proposed b.Y Chermin and
van Krevelenl5• A, represents the untreated coal; P. the plastic
stage; SC, the semi-coke and~, the low temperature gas products
from the primary decomposition.
In these tests chloroform extraction was used to determine
a value proportional to the amount of P produced. The relationship
between the plastic stage and the extract yield seems to be well
established by Dryden and Panchurstl who showed that the residue does
not produce a fluid stage when the extract has been removed.
Procedure
The coal was heated in a pressure tight container as in
previous tests. In order to get reproducibility it was found that
19
20
the heating rate must be carefully controlled. Optimum characteristics
required of the heating process were a rapid rise in temperature to
the maximum temperature for the test and for maintenance of this tempera
ture for a time to produce the greatest extract yield. The rapid
temperature rise was important to minimize reactions before the
maximum temperature was reached and because a rapid temperature rise
usually accentuates the plastic stage and the coking properties of the
coal. It was desirable to hold the maximum temperature for a time
which would produce a maximum yield in order to relate one run to
another and to minimize the effect of errors. After some preliminary
tests the heating pattern that seemed most nearly to satisfy the
preceding conditions was a total heating time of 32 minutes, with the
temperature in the range of the last ten percent of the temperature rise
for approximately forty percent of the heating time. This system was
used for all the heating processes reported in this study.
After heating, the sample was extracted with chloroform and
evaporated as in the previous extraction tests. The yield of extract
was determined and a plot of yield versus temperature was made
(Figure 8). In this plot the temperature used was the mean temperature
for the final ten percent of the temperature rise.
Discussion of Results
Figure 6 shows a plot to determine the order of magnitude of the
activation energy of step 1 of the proposed model. This study utilized
the following relationships:
20
the heating rate must be carefully controlled. Optimum characteristics
required of the heating process were a rapid rise in temperature to
the maximum temperature for the test and for maintenance of this tempera
ture for a time to produce the greatest extract yield. The rapid
temperature rise was important to minimize reactions before the
maximum temperature was reached and because a rapid temperature rise
usually accentuates the plastic stage and the coking properties of the
coal. It was desirable to hold the maximum temperature for a time
loIhich would produce a maximum yield in order to relate one run to
another and to minimize the effect of errors. After some preliminary
tests the heating pattern that seemed most nearly to satisfy the
preceding conditions was a total heating time of 32 minutes, with the
temperature in the range of the last ten percent of the temperature rise
for approximately forty percent of the heating time. This system was
used for all the heating processes reported in this study.
After heating, the sample was extracted with chloroform and
evaporated as in the previous extraction tests. The yield of extract
was determined and a plot of yield versus temperature was made
(Figure 8). In this plot the temperature used was the mean temperature
for the final ten percent of the temperature rise.
Discussion of Results
Figure 6 shows a plot to determine the order of magnitude of the
activation energy of step 1 of the proposed model. This study utilized
the following rela t.ionships:
21
% yield- m kB Heating Time (constant)
-E/RT k = B 1
e
Therefore
ln.kB = ln( constant) J V R T
where k is the reaction rate constant, B and B-̂ are constants, E is
the activation energy for the step involved, R the gas constant and
T the absolute temperature in degrees K,
It follows that on a plot of the rate constant versus l/T
such as Figure °, the slope of the curve must represent E/R. In this
plot the low temperature range has the greatest significance since
it may be assumed that in this temperature range step 2 is negligible.
This plot indicates an activation energy for the initial step of 12 k
cal/mole and 20 k cal/mole for the second step assuming the second
step is controlling beyond the point of maximum yield*
It has been established by van Krevelen, van Heerden and
Huntjens^ that the activation energy for the primary gasification
(E2) is equal to approximately 5>0 k cal/mole for all coals* Chermin
and van Krevelen^ propose that unless E^ is approximately equal to
E2 the coal will not have a plastic stage which will contribute to a
satisfactory coke product* The value of E]_ and E2 determined by these
tests does not appear to agree with previously established values*
A plot similar to Figure 9 using the data of Dryden and Panchurst^
reveals activation energies of the same order of magnitude as for
this stud̂ r*
21
% yield • kB Heating Time (constant)
Therefore
-E/RT e
In.kB = In{constant) ~/RT
where k is the reaction rate constant, B and Bl are constants, E is
the activation energy for the step involved, R the gas constant and
T the absolute temperature in degrees K.
It follows that on a plot of the rate constant versus l/T
such as Figure 9, the slope of the curve must represent E/R. In this
plot the 1m., temperature range has the greatest significance since
it may be assumed that in this temperature range step 2 is negligible.
This plot indicates an activation energy for the initial step of 12 k
ca1/mole and 20 k cal/mole for the second step assuming the second
step is controlling beyond the point of maximum yield.
It has been established by van Kreve1en, van Reerden and
Huntjens14 that the activation energy for the primary gasification
(E2) is eq~l to approxilnate1y 50 k cal/mole for all coals. Chermin
and van Kreve1enlS propose that unless El is approximately equal to
E2 the coal 'Will not have a plastic stage which will contribute to a
satisfactory coke product. The value of El and E2 determined b.1 these
tests does not appear to agree with previously established values.
A plot similar to figure 9 using the data of Dryden and Panchurstl
reveals activation energies of the same order of magnitude as for
this study.
22
Some observations of the physical properties of the coal as
it was removed from the heating process deserve some mention here*
The heated coal appeared much the same as before heating (rather
finely ground) up to the point of maximum yield. At or near the
point of maximum yield the coal sample was caked but not swelled.
Higher temperatures produced a swelled sample with a relatively high
swelling index.
22
Some observations of the physical properties of the coal as
it was removed from the heating process dese~ some mention here.
The heated coal appeared IlRlch the same as before heating (rather
finely ground) up to the point of maximum yield. At or near the
point of maximum yield the coal sample was caked but not swelled.
Higher temperatures produced a swelled sample 'With a relatively high
swelling index.
V. CONCLUSIONS
The use of the infrared spectrophotometry and the extraction
techniques have indicated the following information regarding the
primary decomposition of the coal.
1. The heated coal and the chloroform extract show appreciable
increases in the molecular OH bonds*
2© The chloroform extract yield shows a higher hydrocarbon
content than' the untreated coal. This is particularly significant
since extraction was after heating to temperatures at which some
hydrocarbons leave the material as a gas. The extract must consist
of the lower molecular weight portions of the plastic material of
the coal* This reasoning would imply that the extract is composed
of smaller numbered carbon ring compounds with a large number of
hydrocarbon edge groups.
3o Both the extract and the heated coal show marked changes
in the carbonyl and ether absorbtion region. The results of this
study alon& are not sufficient to discuss these changes except in a
general way. It appears that there is a shift from the carbonyl
bonds toward ether bonds as the coal is heated. A mechanism may be
proposed which includes this change and the increase in the molecular
OH bonds«
k» The extract shows absorbtion at four or more bands for
which there, is no absorbtion in the untreated coal. These absorbtions
23
V. CONCLUSIONS
The use of the infrared spectrophotometry and the extraction
techniques have indicated the following information regarding the
primary decomposition of the coal.
1. The heated coal and the chloroform extract show appreciable
increases in the molecular OR bonds.
20 The chloroform extract yield shows a higher hydrocarbon
content than' the untreated coal. This is particularly significant
since extraction was after heating to temperatures at which same
hydrocarbons leave the material as a gas. The el...-tract must consist
of the lowe~ molecular weight portions of the plastic material of
the coal. This reasoning would imply that the extract is composed
of smaller numbered carbon ring compounds with a large number of
hydrocarbon edge groups.
30 Both the extract and the heated coal show marked changes
in the carbonyl and ether absorbtion region. The results of this
stuqy alon& are not sufficient to discuss these cha-~ges except in a
general way. It appears that there is a shift from the carbonyl
bonds toward ether bonds as the coal is heated. A mechanism may be
proposed which includes this change and the increase in the molecular
OR bonds.
4. The extract shows absorbtion at four or more bands for
which there .. is no absorbtion in the untreated coal. These absorbtions
23
2U
appear to substantiate the belief that there are increased molecular
OH bonds in the heated coalo
There is an eventual disappearance of all or almost all of
the bonds except the C-C bonds as the coal is heated to the coking
temperatures* The final product has a graphitic structure -with the
more fluid coals showing a larger crystallite size.
6* A kinetic study of the initial reaction which results in
a plastic state for the coal indicates activation energies for of
approximately 12 k cal/mole for the first step of the decomposition
process and 20 k cal/mole for the second step of the proposed
mechanism.
It is the author's opinion that the conclusions which may be
drawn from the results of these tests are less important than the
direction of study of which they seem to point. As newer techniques
such as these are developed and as more data become available from
these techniques, a more specific understanding of the primary decom
position of coal will be possible. A good understanding of the
activation energies of the primary decomposition of the coals involved
could change the blending of coals to produce good coke from its
present empirical stage to a much more exact science*
24
appear to substantiate the belief that there are increased molecular
OH bonds in the heated coalo
5. There is an eventual disappearance of all or almost all of
the bonds except the C-C bonds as the coal is heated to the coking
temperatures. The final product has a graphitic structure with the
more fluid coals showing a larger crystallite size.
6. A kinetic study' of the initial reaction which results in
a plastic state for the coal indicates activation energies for of
approximately 12 k cal/mole for the first step of the decomposition
process and 20 k cal/mole for the second step of the proposed
mechanism.
It is the author's opinion that the conclusions which may be
drawn from the results of these tests are less important than the
direction of study of which they seem to point. As newer techniques
such as these are developed and as more data become available from
these techniques, a more specific understanding of the primary decom
position of coal will be possible. A good understanding of the
activation energies of the primary decomposition of the coals involved
could change the blending of coals to produce good coke from its
present empirical stage to a much more exact science.
VI. RECOMMENDATIONS
The studies presented in this report represent only a
beginning in the use of the techniques described to study the primary
decomposition of coal. Infrared absorbtion bands must be more positively
identified, particularly those for the extract yields, for a fuller
understanding of the data obtained by this technique and more data is
necessary for a more thorough kinetic study.
For a better identification of the spectral bands in the
extract, a fractional distillation might be helpful. Gas chromotography
and mass spectrometry could be used to identify the lighter fractions
of the extract.
The fdata which are necessary for a good kinetic study must
include the extract yield as a function of temperature which was
determined in this study and also the extract yield as a function of
time as a constant temperature. Since the plastic stage represented
by the extract yield is decomposing as well as forming, the amount of
the decomposition or weight change of the sample must be obtained also.
The author plans to obtain these required data by the use of vertical
furnace in which weight changes may be recorded as a function of time.
A much more thorough kinetic study should be possible when the
additional data are available.
25
VI. RECOMMENDATIONS
The studies presented in this report represent only a
beginning in the use of the techniques described to study the primary
decomposition of coal. Infrared absorbtion bands must be more positively
identified, particularly those for the extract yields, for a fuller
understanding of the data obtained by this technique and more data is
necessar,y for a more thorough kinetic study.
For a better identification of the spectral bands in the
extract, a fractional distillation might be helpful. Gas chromotography
and mass spectrometry could be used to identify the lighter fractions
of the extract.
The ~da.ta which are necessary for a good kinetic study must
include the extract yield as a function of temperature which was
determined in this study and also the extract yield as a function of
time as a oonstant temperature. Since the plastic stage represented
by the extract yield is decomposing as well as fOrming, the amount of
the decomposition or weight change of the sample must be obtained also.
'ilie author plans to obtain these required data by the use of vertical
furnace in which weight changes may be recorded as a function of time.
A much more thorough kinetic study should be possible when the
additionai data are available.
2.5
APPENDICES APPEmllCES
•Figure 1.
'li. ! I
'i i "1-,· ~ I~: , ,
'j : I ~', i
'I . . , .,
, !
I:
" , '.J : __
If : ~ -; i
I c" . ~ j 0.1 r
.. ,t . j ,J I
0'" 0 .•
.. ll. o. I I ,I. In~ !i' I 0.' 1 i 0.7
loc. 0.8,
0 00' 0 1.0 ~ I .C C.O
I,.~ .! I: I r I.lI I ! II~
W!LVELENGTH IN MICRONS
;Figure 1. DIfferential iDfrared spectrophotometric paltern. Untreated coal In the reference beam; coal beated to the aqfte,/llDJ range In the sample beam.
: I ~ \ 1 .~ r j ! i ~
0.'
0.-
WAVELENGTH IN MICRONS
Figure 2. Differential infrared spectrophotometric pattern to prove technique. Untreated coal from the same sample in both the reference and sample beams.
/
lill mil " .
I' i
Iii
'.. '. ; I' " , .' 1 .1 ,T
I L: i I.: i I 'I II I
II' ,
: I ['
I i
, : ~. .[ ! ;
WAVELENGTH IN MICRON!!
Figure 2. DltferentlallDfrare<l 'spectrophotometric p-artern to prove technique. Untreated coal from tbe same sample in both tba reierence and sample beams.
1
,I
, 'I
: I I
·1 I'
'"I'
I'\) co
Figure 3. Infrared spectrophotometry pattern of coals heated in a closed crucible.
29
2.1 5 4 6 7. 7.5 10
---0 - Furnace at SOOOF, coal sample 1 gram -··-u - In furnace 2 mln~a -·-1 - In furnace 1 minutes ---m - In furnace 3 Dllnutea
- IV - In furnace 5 minutes (max. coal temperature approXimately 4'15"C.
Figure 3. Infrared spectropllotometrlc pattern of coals beated In a closed crucible.
Figure 5. Chloroform extract yield vs. blank,
! !
IJ ~I:
I'
. !
, ",~ ",~
j :
Figure 4. Untreated coal vs. potassium iodide blank ,
1 ' ;
, , ,
.j '
'1"
, ' i
-~: , ~~~~~~~~~ , i
,"
Figure 5. Chloroform extract yield vs. blank.
w o
Figure 7 Extract residue vs. potassium iodide blank
~ . "'I'? , ~ .'.1'" ,~ '?' ': .. " ,,~ . '~ '1" '1" , lIHf " . ':.' : i I I ~ i ' I i II
, Ii, i \,,! _Zl ! I ! , I i I
I I!i: I~i \11 ! :'1
,
I I I I
I . I
. , ! I rl~ ; ,~ I I ! I I I I I i 'I
i II ! I I I II [ I I I I
i I
;' ! I
. , II : I. ! I i ,
!
I I I' I i' 1
,I . I I! I . i I I ,
flml± ! ,
I
mm· I I i II I: I . ! i! ! i
Iffffifi ' . , ... , ,
Figure 6 _. Untreated coal reference beam. Chloroform extract yield sample beam ....
" "I" .. "'. . ~ ."i" . " .:'1". " 'i". ".m "i". '1" ."1" '1" '1" ,
i ~ [ •
,
I . i •
, I , , ,
I
, I
i' I i I . ,
; ; , I I ,
i ,
i , !
, II I ! i , I' , i: ,
! I , , , ,
; I : , , , , ,
, , :
L : , I
. , I ,
, I ,
~ , . I
! I, rIl , . . , .
, , ,
• i , , ,
• i I i I ! II i I ii, I I I ! . i , , '
Ie , . ......
Figure 7.. Extract residue vs. potassium iodide blank.
32
T 1 1 1 1 1 r
200 240 280 320 360 400 440 480 520 T E M P E R A T U R E ° C
Figure 8. Chloroform extraction yield v s . temperature .
31 1 1 1 ' 1 1 1 r
Figure 9. Ln. % extract ion y ie ld v s . 1/T to determine activation e n e r g i e s .
0 ...J UJ
~
z 0 ~ t.)
~ It: ~ )(
UJ
~ °
o
6
5
4
3
2
.5
200
32
520 TEMPERATURE °c
Fi~ure 8. Chloroform ,.'xtractioJ1 :'iield \'5. lenlj)erature.
3.--------.--------.--------r--------~------~--------~-----,
~ 2 ~
~ t.)
ct a: .... ~
Z ...J
1.3
keol Et - 20 mole
o
1.4 1.5 1.7
keol = 12.4 mole
o
1.8
Figure 9. Ln. % extraction yield vs. liT to determine activation energies.
1.9
BIBLIOGRAPHY
1 I.G.C. Dryden and K.X. Panchurst, Fuels, 3U 363, July (1955)
2 M.M. Stimson and M.J. O'Donnell, Jour. Am Chem Soc 7k 1805 (1952)
3 U. Schiedt and H. Reinwein, Z. Naturforsch 76, 270 (1952)
k U. Scheldt, Ibid, 86, 16 (1953)
5 R. A. Friedel and J. A. Queiser, Anal. Chem* 28 No. 1 , 22, Jan. (1956) —
6 J. K. Brown, Chem Soc. Journal of Inst, of Fuels, 28 218, May (1955) —
7 J. K. Brown, and P. B. Hirsch, Nature 175 , Ebr. 5, 229 (1955)
8 J. K.-Brown, Chem. Soc. Journal, pp. 752-757, Jan.-Mar. (1955)
9 Fuels and Combustion Handbook, Johnson and Auth, 1 s t , Edition, McGraw-Hill, New York, (1951)
10 Chemistry of Coal Utilization, Vol. I, John Wiley, (19^5)
1 1 H. H. Storch, Jour, of Inst, of Fuels, 28 151*, April (1955)
12 A. Whitaker, Jour, of Inst, of Fuels, 28 218, May (1955)
13 D. Fitzgerald, Fuel, J35 178 April (1956)
lU D. W. van Krevelen, C# van Heerden and F. J. Huntjens, Fuel, 30 253, (1951)
15 H. A. G. Chermin and D. W. van Krevelen, Fuel, 36 85, Jan. (1957)
16 W. A. KLnkby, J. R. A. Lakey and E. J. Sargent, Fuels, 33, 1*80. (195U)
33
BIBLIOGRAPHY
1 IoG.Co Dryden and K.X. Panchurst, Fuels, 34 363, July (1955)
2 MoM. stimson and M.J. O'Donnell, Jour. Am Chern Soc. 74 1805 (1952) --
3 U. Schiedt and H. Reinwein, Z. Naturforsch 76, 270 (1952)
4 U. Scheidt, Ibid, 86, 16 (1953)
5 R. A. Friedel and J. A. Queiser, Anal. Chem. 28 No.1, 22, Jan. (1956)
6 J. K. Brown, Chern Soc. Journal of Inst. of'Fuels, 28 218, May (1955)
7 J. K. Brown, and P. B. Hirsch, Nature 175, Ebr. 5, 229 (1955)
8 J. K. -'Brown, Chern. Soc. Journal, pp. 752-757, Jano-¥...ar. (1955)
9 Fuels and Combustion Handbook, Johnson and Auth, 1st. Edition, McGraw-Hill, New York, (1951)
10 Chemistry of Coal Utilization, Vol. I, John Wiley, (1945)
11 H. H. Storch, Jour. of Inst. of Fuels, ~ 154, April (1955)
12 A. Whitaker, Jour. of Inst. of Fuels, 28 218, May (1955)
13 D. Fitzgerald, Fuel, ~ 178 April (1956)
14 D. W. 'van Krevelen, C. van Heerden and F. J. Huntjens, Fuel, .lQ. 253, (1951)
15 H. A. G. Chermin and D. W. van Krevelen, Fuel, .36 85, Jan. (1957)
16 W. A. Kinkby, J. R. A. Lakey and R. J. Sargent, Fuels, ll, 480, (1954)
33
A STUDY OF THE PRIMARY DECOMPOSITION OF COAL BY INFRARED
SPECTROPHOTOMETRY AND BY CHLOROFORM EXTRACTION
by
Raymond Virgil Smith
An abstract of a thesis submitted to the faculty of the University of Utah in partial fulfillment of the requirements for the degree of
Master of Science
Approved by the faculty committee in
August, 1957
Dr. George Richard Hill, Chairman, Supervisory Committee
Department of Fuel Technology
University of Utah 1957
A STUDY OF THE PRIMARY DECOMPOSITION OF COAL BY INFRARED
SPECTROPHOTOMETRY AND BY CHLOROFORM EXTRACTION
by
Raymond Virgil Sm:i. th
An abstract of a thesis submitted to the faculty of the University of Utah in partial fulfillment of the requirements for the degree of
Master of Science
Approved by the faculty committee in
August, 1957
Dr. George Richard Hill, Chairman, Supervisory Committee
Department of Fuel Technology
Uni versi ty of Utah 1957
ABSTRACT
For the differential I-R technique described in this paper,
two specimens of coal were prepared for runs in the infrared double
beam spectrometer. One specimen was prepared for the coal as
received from the mine and this was placed in the reference beam.
The other sample was prepared from coal which had been heated in
a pressure tight container to the softening temperature. This was
placed in the sample beam. The differential infrared spectrometer
pattern thus obtained enables one to observe the changes in the
infrared range.
In the second phase of these tests the coal was heated to
temperatures in the plastic range and then extracted with chloroform.
The extract yield was run differentially versus the untreated coal
in the infrared spectrophotometer. These tests indicated different
band intensities than the untreated coal and also revealed some
absorbtion bands which did not occur in the original coal or in the
residue extract from the absorbtion process.
The extract yield data was also used for a kinetic study of
the coal!s primary decomposition. Activation energies thus obtained
for the solid to plastic step of the reaction appear to be of the
general order of magnitude of 20 to 30 k.cal/mole*
ii
ABSTRACT
For the differential I-R technique described in this paper"
two specimens of coal were prepared for runs in the infrared double
beam spectrometer. One specimen was prepared for the coal as
recei ved from the mine and this was placed in the reference beam.
The other sample was prepared from coal which had been heated in
a pressure tight container to the softening temperature. This was
placed in the sample beam. The differential infrared spectrometer
pattern thus obtained enables one to observe the changes in the
infrared range.
In the second phase of these tests the coal was heated to
temperatures in the plastic range and. then extracted with chloroform.
The extract yield was run differentially versus the untreated coal
in the infrared spectrophotometer. These tests indicated different
band intensities than the untreated coal and also revealed some
absorbtion bands which did not occur in the original coal or in the
residue extract from the absorbtion process.
The extract yield data was also used for a kinetic study of
the coal r s primary de compo si tion. Activation energies thus obtained
for the solid to plastic step of the reaction appear to be of the
general order of magnitude of 20 to 30 k.cal/mole.
i1